Early and habitual alcohol use is prodromal to alcohol dependence. Multiple factors including genetics and environment contribute to early and habitual drinking. Numerous studies have suggested cerebral structures and functions as an intermediate phenotype of alcohol misuse and vulnerability to alcohol dependence. On the other hand, while there is abundant information on how cerebral structures and functions are altered as a result of chronic alcohol use, little is known about these influences in non-dependent early and habitual drinking. In particular, what are the neural correlates that would predict increase in problem drinking and the quantity and frequency of alcohol use in non-dependent individuals? To address this question, we propose to recruit a large sample of early alcohol drinkers for magnetic resonance imaging (MRI) studies and clinical follow-up. Volunteers will undergo detailed medical and psychiatric assessments including their drinking behaviors, participate in MRI studies to evaluate cerebral structures (voxel-based morphometry or VBM), white matter integrity (diffusion tensor imaging or DTI), and functions of the component processes of cognitive control and alcohol cue reactivity (functional MRI or fMRI). We will engage all participants in a two-year clinical follow-up with tri-monthly assessments of their drinking behaviors. Head hair ethyl glucuronide will be quantified at baseline and every other follow-up, as a complementary index of alcohol consumption. For each of these neural measures, we will examine how they predict changes in drinking behavior, both independently and synergistically. Our overarching goal is to identify cerebral endophenotypes that best describe non-dependent early, habitual drinking and predict changes in problem drinking as well as the frequency and quantity of drinking. The potential findings would characterize the circuit biomarkers of early habitual drinking at a level of details that are critically needed and yet not currently available. Most importantly, these neural markers would help identify and facilitate early intervention for individuals at risk of developing heavy drinking and alcohol dependence.